System and method for low latency node local scheduling in distributed resource management

Abstract

A system for allocation of resources and processing jobs within a distributed system includes a processor and a memory coupled to the processor. The memory includes at least one process and at least one resource allocator. The process is adapted for processing jobs within a distributed system which receives jobs to be processed. The resource allocator is communicably coupled with at least one process, and is adapted to generate one or more sub-processes within a limit of one or more resources allocated to the process for processing jobs.

What is claimed is: 
     
       1. A device, comprising:
 a processor; and 
 a memory coupled to the processor and storing instructions that, when executed by the processor, cause the processor to be configured to:
 communicate with a resource master to allocate node resources on a node of a distributed system; 
 generate a master container after allocation of the node resources; 
 allocate container resources for the master container from the node resources, wherein the container resources comprise resources for running the master container and buffer resources, and wherein the buffer resources comprise a total amount of memory for processing job requests in the device; 
 receive the job requests; 
 generate sub-containers within the master container, wherein each of the sub-containers comprises a portion of the total amount of memory for processing the job requests in the device; 
 allocate, by the processor for each of the job requests, one of the sub-containers to process the job request; 
 reallocate, by the processor for each of the job requests, the one of the sub-containers to another one of the job requests when the portion of the total amount of memory is not being utilized by the job request; 
 reallocate, by the processor, the portion of the total amount of memory from the one of the sub-containers to a higher priority job request by pausing the one of the sub-containers when the higher priority job request is received by the device; and 
 resize, by the processor, sizes of the master container and the sub-containers based on resource usage of the device. 
 
 
     
     
       2. The device of  claim 1 , wherein the instructions further cause the processor to be configured to track the allocation of the node resources. 
     
     
       3. The device of  claim 1 , wherein the instructions further cause the processor to be configured to determine whether the allocation of the buffer resources for each of the sub-containers is within a limitation preset by the master container. 
     
     
       4. The device of  claim 1 , wherein the instructions further cause the processor to be configured to resize the resources for each of the sub-containers. 
     
     
       5. The device of  claim 1 , wherein the instructions further cause the processor to be configured to pause each of the sub-containers. 
     
     
       6. The device of  claim 1 , wherein the master container is configured to coordinate with each of the sub-containers. 
     
     
       7. The device of  claim 1 , wherein the job request comprises an ad-hoc job request. 
     
     
       8. The device of  claim 1 , wherein a size of each of the sub-containers is within resource limits of the master container. 
     
     
       9. A method implemented by a device, the method comprising:
 communicating with a resource master for allocation of node resources; 
 generating a master container after the allocation of the node resources by the resource master; 
 allocating container resources for the master container from the node resources, wherein the container resources comprise resources for running the master container and buffer resources, and wherein the buffer resources comprise a total amount of memory for processing job requests in the device; 
 receiving the job requests; 
 generating sub-containers within the master container, wherein each of the sub-containers comprises a portion of the total amount of memory for processing the job requests in the device; 
 allocating, for each of the job requests, one of the sub-containers to the job request to process the job request; 
 reallocating, for each of the job requests, the one of the sub-containers to another one of the job requests when the portion of the total amount of memory is not being utilized by the job request; 
 reallocating the portion of the total amount of memory from the one of the sub-containers to a higher priority job request by pausing the one of the sub-containers when the higher priority job request is received by the device; and 
 resizing sizes of the master container and the sub-containers based on resource usage of the device. 
 
     
     
       10. The method of  claim 9 , further comprising tracking of the allocation of the node resources. 
     
     
       11. The method of  claim 9 , further comprising determining whether the allocation of the buffer resources for each of the sub-containers is within a limitation preset by the master container. 
     
     
       12. The method of  claim 9 , further comprising resizing the resources for each of the sub-containers. 
     
     
       13. The method of  claim 9 , further comprising pausing each of the sub-containers. 
     
     
       14. The method of  claim 9 , further comprising coordinating the master container with each of the sub-containers. 
     
     
       15. The method of  claim 9 , wherein the job request comprises an ad-hoc job request. 
     
     
       16. The method of  claim 9 , wherein a size of each of the sub-containers is within resource limits of the master container. 
     
     
       17. A computer program product comprising computer-executable instructions stored on a non-transitory computer-readable medium that, when executed by a processor, cause a computing device to:
 communicate with a resource master to allocate node resources; 
 generate a master container after the allocation of the node resources by the resource master; 
 allocate container resources for the master container from the node resources, wherein the container resources comprise resources for running the master container and buffer resources, and wherein the buffer resources comprise a total amount of memory for processing job requests in the computing device; 
 receive the job requests; 
 generate sub-containers within the master container, wherein each of the sub-containers comprises a portion of the total amount of memory for processing the job requests in the computing device; 
 allocate, for each of the job requests, one of the sub-containers to process the job request; 
 reallocate, for each of the job requests, the one of the sub-containers to another one of the job requests when the portion of the total amount of memory is not being utilized by the job request; 
 reallocate the portion of the total amount of memory from the one of the sub-containers to a higher priority job request by pausing the one of the sub-containers when the higher priority job request is received by the computing device; 
 resize sizes of the master container and the sub-containers based on resource usage of the computing device. 
 
     
     
       18. The computer program product of  claim 17 , wherein the instructions further cause the computing device to determine whether the allocation of the buffer resources for each of the sub-containers is within a limitation preset by the master container. 
     
     
       19. The computer program product of  claim 17 , wherein the instructions further cause the computing device to coordinate the master container with each of the sub-containers. 
     
     
       20. The computer program product of  claim 17 , wherein a size of each of the sub-containers is within resource limits of the master container.